Recommendations of the Expert Panel to Define Removal Rates for Urban Nutrient Management

Transcription

1 Recommendations of the Expert Panel to Define Removal Rates for Urban Nutrient Management FINAL REPORT VERSION 2.0 Submitted by: Marc Aveni, Karl Berger, Jonathan Champion, Gary Felton, Mike Goatley, William Keeling, Neely Law and Stuart Schwartz submitted to: Urban Stormwater Work Group Chesapeake Bay Partnership February 3, 2013 Prepared by: Tom Schueler and Cecilia Lane Chesapeake Stormwater Network

2 Table of Contents Page Executive Summary 4 Section 1 Charge and Membership of the Panel 6 Section 2 Definitions and Qualifying Conditions 8 Section 3 Turf and Fertilization Behavior in Chesapeake Bay 12 2 Estimating urban pervious land and turf cover in the Bay watershed 12 Status of state phosphorus fertilizer legislation 14 Trends in non-farm fertilizer sales in the Bay watershed 15 Derivation of the original CBP-approved rate for UNM 18 How CBWM simulates nutrient loads from pervious areas 18 Section 4 Review of the Available Science 21 Review of phosphorus dynamics on urban lawns 21 Review of nitrogen dynamics on urban lawns 22 High risk factors for nutrient export 25 Scientific justification for core UNM practices 28 Regional studies on effect of P fertilizer restrictions 32 Summary of homeowner fertilizer behaviors 33 Effect of outreach in changing fertilizer behavior 34 Section 5 The Recommended Credits and Rates 38 State-wide P reduction credit for urban pervious land 38 State-wide N reduction credit for urban pervious land 40 N and P reduction credits for UNM practices 42 Program credit for the alternative outreach option 44 No credit for passive outreach 44 Section 6 Accountability Mechanisms 45 Verification for state-wide reduction credits 45 Verification for UNM practices 47 Verification of the alternative outreach option 48 Reducing the potential for double counting 49 Section 7 Future Research and Management Needs 49 Justification of the recommendations 49 UNM capacity, communication and delivery issues 50 Proposed CBWM refinements 51 Priority research recommendations to fill management gaps 52

3 References Cited 53 Appendix A Mass Balance Check on UNM Rates Appendix B Nutrient Management on Public Lands: A Review Appendix C Sample Urban Nutrient Management Plan Appendix D Consolidated Meeting Minutes of the Panel Appendix E Conformity with BMP Review Protocol List of common acronyms used throughout the text: BMP CAST CBP CBWM DIY GIS HOA MS4 Rv RT VM Sf SRP TMDL TN or N TP or P TSS UNM WIN WIP WQGIT WTM Best Management Practice Chesapeake Assessment Scenario Tool Chesapeake Bay Program Chesapeake Bay Watershed Model Do it Yourself Geographic Information Systems Homeowner Association Municipal Separate Storm Sewer System Runoff Coefficient Reporting, Tracking, Verification and Monitoring Square feet Soluble Reactive Phosphorus Total Maximum Daily Load Total Nitrogen Total Phosphorus Total Suspended Solids Urban Nutrient Management Water Insoluble Nitrogen Watershed Implementation Plan Water Quality Goal Implementation Team Watershed Treatment Model 3

4 4 Summary of Panel Recommendations More than 3.5 million acres of urban pervious lands exist in the Bay watershed, comprising nearly 10% of its total area. This diverse category of land cover includes both fertilized and un-fertilized turf and is managed in many different ways. Bay states have collectively targeted more than 45% of the pervious land for the application of urban nutrient management (UNM) practices to help achieve load nutrient reductions to meet the Bay TMDL by The Panel discarded the existing CBP-approved definition of UNM as being too ambiguous and also concluded that the corresponding removal rates for UNM were not technically justified. The Panel then reviewed more than 200 research studies and reports to understand turf grass N and P dynamics, homeowner fertilization behaviors, the effects of P fertilizer restrictions in watersheds outside of the Bay and the effect of various outreach campaigns to change those behaviors. The Panel also examined historic and recent trends in fertilizer sales across the watershed and confirmed the general adequacy of the technical assumptions for fertilizer inputs to pervious lands in the CBWM. The literature review supported the contention that most turf grass is highly retentive of applied N, but may still export some particulate organic N regardless of whether a lawn is fertilized or not. The Panel identified 11 site-based factors associated with a high risk of N and P export, such as soils, slope, terrain, age and lawn care practice. These sitebased factors led the Panel to define ten core lawn care practices that minimize the risk of N and P export, which collectively define the UNM practice. Based on the science and best professional judgment, the Panel recommends three types of nutrient reduction credits. The first is an automatic state-wide P reduction credit starting in 2013 that reflects declines in P fertilizer application rates due to recent state phosphorus fertilizer legislation and the gradual industry phase out of P in fertilizer products. The exact reduction varies by state, but is about 25% for states that have adopted legislation and 20% for those that have not. The automatic credit expires in three years, and will be replaced by a more verifiable and variable credit based on declines in unit area P application rates derived from improved non-farm fertilizer sales statistics. States may also be eligible for a state-wide N reduction credit in 2014 if they can document declines in unit N fertilizer applications relative to the current application rate benchmark employed in the CBWM. The second credit is a removal rate for the acreage of pervious land covered by qualifying UNM practices, based on the site risk for N and P export. For low risk lawns, the UNM load reductions for TN and TP are 3 and 6% respectively. The load reductions increase when UNM practices are applied to high risk lawns (20% TN, 10% TP). The third credit is granted to communities that experiment with alternative outreach programs to improve UNM delivery and perform surveys and monitoring to advance understanding of how to change existing fertilizer behaviors. Communities that make

5 the investment in alternative outreach programs are provided a one-time credit of a 3% reduction in the TN load from the pervious land in the geographic area in which their outreach program is targeted. Summary of Urban Fertilizer Management Credits for Phosphorus and Nitrogen Nutrient Statewide with P fertilizer legislation Statewide without P fertilizer legislation Urban Nutrient Management UNM 1, 2 Alternative Outreach Option Phosphorus 25% 20% Low risk: 3% High risk: 10% Blended: 4.5% Alternative: 0% Passive: 0% Notes & Conditions of Credit Effective 2013 for 3 years. In 2016, need to show reduction in P using two years of fertilizer sales data Need to survey highrisk every 5 years; Renew UNM every 3 years Study needed to quantify outcomes and be statistically relevant. Nitrogen Notes & Conditions of Credit 3% for every 10% decrease in N urban fertilizer input from CBWM benchmark Effective 2014, need to show N reduction using two consecutive years sales data Low risk: 6% High risk: 20% Blended: 9% Need to survey highrisk every 5 years; Renew UNM every 3 years Alternative: 3% Passive: 0% Study needed to quantify outcomes and be statistically relevant. The Panel developed methods for reporting, tracking and verifying all three credits to ensure the UNM practices achieve their intended pollutant reduction. The Panel acknowledged that there are still many unknowns when it comes to the UNM practice, and adopted an adaptive management approach as it developed its recommendations. The Panel also recommended improvements to the CBWM model and priority research projects that could improve confidence in its representation of UNM. Lastly, the Panel recommended several ways to improve Bay-wide communication of the UNM message, and improve the capacity to deliver UNM practices to meet the future demand for this practice. 5

6 Section 1 Charge and Membership of the Panel Urban Nutrient Management Expert Panel Panelist Affiliation Jonathan Champion District Department of the Environment Karl Berger Metropolitan Washington Council of Governments Dr. Stu Schwartz University of Maryland, Baltimore County William Keeling Virginia Department of Conservation and Recreation Dr. Gary Felton University of Maryland, College Park Dr. Neely Law Center for Watershed Protection Marc Aveni Prince William County Department of Public Works Dr. Mike Goatley Virginia Tech Tom Schueler Chesapeake Stormwater Network (panel facilitator) Technical support by Jeremy Hanson (CRC), Molly Harrington (CRC), Gary Shenk (EPA CBPO, Jeff Sweeney (EPA CBPO) and Mark Sievers (TetraTech) is gratefully appreciated The initial charge of the Panel was to review all of the available science on the nutrient removal rates associated with four kinds of nutrient management practices applied to urban pervious areas. 1. Automatic credit for State-wide phosphorus fertilizer legislation 2. Possible credit for jurisdictions without phosphorus fertilizer legislation that reflect industry phase out of P in fertilizer products 3. Proper fertilizer application on privately and publicly owned turf (i.e., Urban Nutrient Management) 4. Local outreach campaigns to reduce fertilization frequency on privately-owned turf The Panel was specifically requested to assess: Current CBWM land use data for urban pervious areas and recommend the most probable splits for turf management status (i.e., fertilized, un-fertilized, and over-fertilized), based on homeowner surveys, sales data, land cover and other metrics. Available literature on the nutrient and sediment loading rates associated with fertilized, un-fertilized and over-fertilized turf, accounting for regional and terrain differences. Current CBWM modeling assumptions to simulate the impact of reduced P applications to pervious areas as a result of adoption of state-wide phosphorus fertilizer legislation. Specific definitions for each class of nutrient management practices and the qualifying conditions and rationale under which a jurisdiction can receive a nutrient reduction credit. 6

7 Whether the existing CBP approved nutrient load reduction rates for urban nutrient management practices developed in 2003 are still reliable, recommend minimum local outreach and education program requirements needed to qualify for them, and how jurisdictions will be able to certify the acreage where the practices are implemented. Extent of fertilizer applications on public lands, and recommend the minimum changes in local landscaping, purchasing and contracting policies in order to reduce the frequency of un-needed fertilizer applications. The Panel may also recommend procedures to evaluate better nutrient management practices on local, state and federal lands. What, if any, nutrient credits can be provided by outreach campaigns to change homeowner behavior from lawn fertilization to non-fertilization (as well as any increase or decrease in sediment delivery). If such a credit is proposed, the Panel will need to define the metrics that communities will need to measure to certify that the change in fertilizer behavior actually takes place. The proper units to report urban nutrient management (UNM) implementation to receive credit in the Chesapeake Bay Watershed Model The Panel confined its efforts to managed urban turf (including golf courses) and did not address turf farms, highway medians or temporary/permanent vegetative stabilization at construction sites. Beyond this specific charge, the Panel was asked to: Determine whether to recommend that an interim BMP rate be established for one or more classes of urban nutrient management practices prior to the conclusion of the panel for WIP planning purposes Recommend procedures to report, track and verify that urban nutrient management practices are actually being implemented on the ground Critically analyze any unintended consequences associated with the nutrient management credit and any potential for double or over-counting of the credit While conducting its review, the Panel followed the procedures and process outlined in the WQGIT BMP review protocol (WQGIT, 2010). The process begins with BMP expert panels that evaluate existing research and make initial recommendations on removal rates. These, in turn, are reviewed by the Urban Stormwater Workgroup and the Watershed Technical Workgroup to ensure they are accurate and consistent with the Chesapeake Bay Watershed Model (CBWM) framework. Appendix D documents the process by which the Panel reached consensus, in the form of a series meeting minutes that summarize their deliberations. Appendix E documents how the Panel satisfied the review criteria established in the BMP review protocol. 7

8 Section 2 Definitions and Qualifying Conditions The Panel agreed that the UNM practice has been ambiguously defined in the past in the context of the CBWM, and therefore expended a great deal of effort to come up with stronger definitions and qualifying conditions so that any reduction credits could be accurately reported, tracked and verified. With this in mind, the Panel came to consensus on the following definitions: Pervious Land: This term is used to describe urban and suburban land that is not impervious in the Chesapeake Bay Watershed Model (CBWM). This land use category predominately includes residential lawns, but may also include landscaping, gardens, parks, rights of way, vacant lots and open areas. Pervious land may also include a limited amount of forest canopy. Pervious lands are subject to different management regimes including just periodic mowing all the way up to the intensive maintenance of a golf course. In the context of the CBWM, fertilizer inputs to pervious areas are currently represented by a single weighted average for both fertilized and un-fertilized pervious areas (i.e., all pervious areas receive fertilizer input). Turf (aka lawns, turf grass, turf cover): In the context of this report, the term turf refers primarily to pervious areas that are managed to attain dense grass cover, which may involve one or more of the following: fertilization, irrigation, weed control, and other turf management practices. Conservation Landscaping: Creation of mulched beds that contain plantings of perennial herbaceous plants, shrubs and small trees that retain rainfall and adsorb runoff from adjacent turf or impervious cover. Native plants are preferred, but ornamental plants are acceptable if they are adapted to regional climates and are not invasive spreaders. High Risk Export Factors: These are defined as pervious areas that are subject to one or more of the following risk factors: 1. Currently over-fertilized beyond state or extension recommendations 2. P-saturated soils as determined by a soil P test 3. Newly established turf (i.e., less than three years old) 4. Steep slopes 5. Exposed soil 6. High water table 7. Over-irrigated lawns 8. Soils that are sandy, shallow, compacted or have low water holding capacity 9. High use areas (e.g., athletic fields, golf courses) 10. Adjacent to stream, river or Bay 11. Karst terrain More specific operational definitions of each risk factor are described in Section

9 Statewide Phosphorus Reduction Credit for Pervious Land: This load reduction credit is determined for each state to reflect the impact of phosphorus fertilizer legislation and/or the gradual P phase out in the market. The automatic credit is initially based on the assumed annual P fertilizer inputs for pervious land in the most recent version of the CBWM. In 2016, however, the state credit will be adjusted upward or downward, based on state-reported trends in the P content of non-farm fertilizer sales data. Statewide Nitrogen Reduction Credit for Pervious Land: This load reduction credit is determined for each state to reflect the expected decline in N fertilizer sales over time. The credit will be initially based on each state's 2014 N fertilizer inputs, relative to the current CBWM assumption of 43 lbs/ac/year for pervious land, and will only be granted if states can document a downward trend in the N content of non-farm fertilizer sales data. The magnitude of the credit will be determined by changing N fertilization inputs in the CBWM. This credit will also be subject to biennial verification. Urban Nutrient Management: is defined as identifying how the major plant nutrients (nitrogen, phosphorus, and potassium) are to be annually managed for expected turf and landscape plants and for the protection of water quality. A nutrient management plan is a written site specific plan which addresses these issues. The goal of an urban or turf and landscape nutrient management planning is to minimize adverse environmental effects, primarily upon water quality, and avoid unnecessary nutrient applications. It should be recognized that some level of nutrient loss to surface and groundwater will occur even by following the recommendations in a nutrient management plan, however, these losses should be lower than would occur without nutrient management (VCE, 2011). Table 1 outlines some of the required elements of an urban nutrient management plan in Virginia. In addition, a sample copy of UNM plan is provided in Appendix C. Core UNM Practices. The Panel concluded that the ten lawn care practices outlined in Section 4.4 and summarized in Table 2 constitute effective UNM practice in the Chesapeake Bay. These ten practices should be reinforced in the core outreach message communicated to the public, and as many practices as might apply to a site should be incorporated into a UNM plan or homeowner pledge. It is recognized that some states may modify the individual lawn care practices to meet their own unique terrain and conditions, as long as they document the nutrient reduction benefit. Phosphorus Fertilizer Legislation: Refers to the passage and implementation of state legislation to restrict the P content in lawn maintenance fertilizer and require or recommend other nutrient management practices on urban turf. As described in Section 3.2, each of the three Bay states has taken different approaches in their legislation. Some fertilizer P application may still be allowed in several Bay states, so the Panel has avoided the term P-ban in this report, except when reviewing the impact of local ordinances enacted in non-bay states. 9

10 Table 1 Common Components of an Urban Nutrient Management Plan in VA 1. Use tables in VA DCR (2005) and soil test information to develop plant nutrient recommendations 2. Calculate phosphorus application rates based on soil test. 3. Know when phosphorus applications are not allowed based on soil test phosphorus saturation level. 4. Understand specific nitrogen management criteria when dealing with environmentally sensitive sites as related to various nitrogen sources and plants 5. Develop a schedule for the timing and placement of fertilizers 6. Develop an integrated nutrient balance sheet for all nutrient sources, application rates and timings 7. Understand issues to address in a plan narrative 8. Determine hydrologic unit code from Virginia National Watershed Boundary Dataset maps 9. Generate appropriate maps to: a. show site and boundaries where nutrients will be applied, b. delineate management areas and indicate size in acres or square feet, environmentally sensitive areas, c. setback areas for application of organic materials. 10. Identify character of disturbed, imported or manufactured soils and determine appropriate nutrient management related management considerations 11. Determine how to define management areas as a function of use or vegetation type and how that impacts nutrient application 12. Determine available nutrient application rates from a wastewater nutrient analysis and the amount of water applied (in the case of wastewater reuse) 13. Determine acceptable periods of nitrogen application for various turf grass types based on location in Virginia and characteristics of the fertilizer to be applied 14. Selection and management of de-icing materials to reduce water quality impact 15. Employ stormwater management principles to reduce runoff pollution Source: Adapted from VA DCR (2005) Table 2 Core Urban Nutrient Management Practices for the Chesapeake Bay 1 Consult with the local extension service, master gardener or certified applicator to get technical assistance to develop an effective urban nutrient management plan for the property. 2 Maintain a dense vegetative cover of turf grass and/or conservation landscaping to reduce runoff, prevent erosion, and retain nutrients 3 Choose not to fertilize, OR adopt a reduce rate/monitor approach OR the small fertilizer dose approach. 4 Retain clippings and mulched leaves on the yard and keep them out of streets and storm drains 5 Do not apply fertilizers before spring green up or after grass becomes dormant 6 Maximize use of slow release N fertilizer during the active growing season 7 Set mower height at 3 inches or taller 8 Immediately sweep off any fertilizer that lands on a paved surface 9 Do not apply fertilizer within 15 to 20 feet of a water feature (depending on applicable state regulations) and manage this zone as a perennial planting, meadow, grass buffer or a forested buffer 10 Employ lawn practices to increase soil porosity and infiltration capability, especially along portions of the lawn that convey or treat stormwater runoff. 10

11 Nitrogen Fertilization Legislation (Maryland Only). This refers to state legislation or regulations that: (a) limits the N content and establishes minimum slow release content for DIY fertilizer products sold in retail outlets (b) sets an upper limit on the maximum amount of N fertilizer that commercial applicators can apply in any one application (0.9 lbs/acre/year) (c) prohibits application on paved surfaces, water features, or during the dormant season, and, (d) has verifiable procedures for commercial applicator training, certification, and application record-keeping, including fines for non-compliance. Maryland's lawn fertilizer legislation is currently the only Bay state that meets criteria (a) - (d), as outlined in MDA (2013). As a result, the acreage of pervious land serviced by commercial applicators that meet the core UNM practices is eligible for a nitrogen credit, as long they can be verified as conforming with the new regulations. Maryland may also receive a smaller nitrogen credit for the acreage of home lawns managed by doit-yourselfer, that are directly influenced by its new retail sales and labeling requirements under the new regulations. The method used to define the N credit is explained in Section 5.3. The Maryland Nitrogen UNM Credit is subject to the training, certification, record keeping and verification procedures outlined in Section 6.3. UNM Planning Agency: This refers to the specific agency in a community that has authority and/or qualifications to assess a property and prepare a verifiable UNM plan. In most states, the UNM planning agency may be the State Cooperative Extension Service, Soil and Water Conservation District, State Agency, or a Local Agency. In some cases, support may be provided by Master Gardeners, a watershed stewards academy, local watershed groups or landscape contractors associations. Each Bay state may specifically define which agency(s) are responsible for UNM plans in their state (e.g., Virginia). Qualifying Urban Nutrient Management Plan. The basic reporting unit for the practice is the acreage of written UNM plans or applicator certifications that contain the applicable lawn care practices specified in Table 2, and are subject to verification. Homeowner UNM Pledge: This is a shorter version of a UNM plan in which an individual homeowner submits a written pledge to implement the applicable UNM practices on their lawn, after an on-site visit from a trained professional to assess risk factors and test soils. The nutrient reduction credit for homeowner pledges is less than for lawns that have a qualified UNM plan, and is limited to no more than the low risk UNM credit for both TN and TP. Each Bay state will choose whether homeowner pledges are an allowable UNM delivery option within their jurisdiction. 11

12 Trained UNM Expert: An individual with the requisite training and experience to prepare UNM plans in their jurisdiction. Several Bay states have established voluntary or mandatory training programs to certify UNM experts. Active Outreach Program. This retail outreach effort is designed to directly interact with individual fertilizer applicators to adopt the core UNM practices, along with other Bay friendly landscaping practices. The outreach effort may be targeted to properties with known high risk factors or be applied across the community such that higher credits are granted for outreach that focuses on high risk turf grass. The product of this strategy is a verifiable UNM plan or pledge whereby an individual homeowner, lawn care company, HOA, business, institutional or public landowner commits to the applicable lawn care practices that apply to their turf. Alternative Outreach Option: States, local governments or other entities have the option to create alternative outreach programs that motivate property owners or commercial applicators to implement the core UNM practices. Examples might include: 12 Use of local or regional social media/marketing campaigns to deliver the core message to a geographically defined target population, and encourage them to contact a UNM planning agency for their property. Mapping of high risk export factors within a community to target outreach efforts to increase the number of UNM plans prepared for these types of properties. Targeted training to commercial applicators, lawn care companies, landscape contractors and property managers to encourage them to adopt UNM practices. Point of sale outreach at retail sales outlets to encourage UNM practices. Other alternative outreach efforts, as approved by state and/or extension service. The acreage of UNM plans and/or pledges is still the metric by which alternative outreach methods will be credited. However, the agency that sponsors the alternative outreach effort will also be eligible for a three year N reduction credit equivalent to 3% of the N load generated for the total pervious area within the geographical area being targeted. The program credit is contingent upon the implementation of a before and after survey to measure actual changes in fertilization behavior outcomes in the target community. After the three years, the alternative outreach program credit will be adjusted upwards or downwards based on the actual survey results and/or UNM plans written. Alternative outreach efforts may need to be repeated and verified to be eligible for program credit on an ongoing basis. The measurement and verification procedures for the credit are outlined in Section 6.3. Passive Outreach. This outreach effort uses non-targeted and wholesale methods to get the UNM message out to the public. Wholesale methods have been traditionally used by MS4 communities to fulfill their stormwater education requirements. These methods may include providing information about proper fertilization on a local web pages, newsletters, kiosks, fact sheets and handbooks. The Panel could find no evidence to support a nutrient reduction credit for passive outreach.

13 Section 3 Background on Turf and Fertilization in the Chesapeake Bay 3.1 Estimating Urban Pervious Area and Turf Cover in the Bay Watershed Until recently, the extent of turf cover associated with urban, suburban and exurban land development in the watershed has been poorly understood. The acreage of turf cover has steadily increased in the Bay watershed over the last four decades as farms and forests have been converted into new development (Schueler, 2010). With new development, small parcels of turf cover are interspersed within a broader mosaic of land use that make it a challenge to characterize (Claggett et al, 2011). Turf cover may also be hidden by tree canopy or confused with pasture in exurban areas. As a result, turf cover within highway rights of way, parks, golf courses, airports, residential lots, cemeteries, schools, churches, hobby farms and institutions may not always be well represented in urban land cover classifications. Consequently, turf cover has been hard to detect directly from satellite imagery, aerial photography or GIS analysis. Recent work by Claggett et al (2011) and Schueler (2010), however, have developed updated estimates of the extent of pervious lands in the Chesapeake Bay using multiple methods. The studies independently calculated that pervious land covers about 3.8 million acres in the Bay watershed, or just less than 10% of the total watershed area. To put this in perspective, turf cover is now equivalent to the area devoted to row crops (corn, soybeans, wheat) in the Bay watershed. The estimated acreage of turf cover in each Bay state is provided in Table 3, and the general distribution of turf cover is portrayed in Figure 1. Based on these new methods, the acreage of pervious land simulated in the CBWM has increased by more than a million acres from Version 4 to Version The extent of turf cover predicted by the methods of Claggett et al (2011) for the CBWM showed reasonable agreement with higher resolution estimates of turf cover for Baltimore County, MD, and further testing is now occurring in other Bay counties (Claggett, 2012). 13

14 Figure 1: Distribution of Turf Cover in the Chesapeake Watershed (Schueler, 2010). Table 3 Estimated Distribution of Urban Pervious Land in the CBWM 5.3.2, By Bay State Urban Pervious Area 1 Acres State Delaware 36,481 District of Columbia 17,206 Maryland 990,291 New York 170,716 Pennsylvania 1,052,558 Virginia 1,195,567 West Virginia 88,218 TOTAL 3,551,037 1 Acres of Urban Pervious Area in Version of Chesapeake Bay Watershed Model About 60 to 80% of pervious land area is associated with residential lawns, depending on the state and reporting era. A summary of these studies can be found in Table 4. More detail on what is known about current homeowner practices on turf can be found in Section

15 Approximately 10 to 15% of pervious land is managed by commercial or institutional land uses. In most cases, they utilize landscape contractors or their own maintenance crews to manage them. The Panel could find very little information on the current UNM practices for this category of pervious land. About 15% to 20% of pervious land is managed by public agencies, in the form of road right of ways, municipal open space, schools and parks. A more detailed discussion of current UNM practices and policies for public turf can be found in Appendix B. Table 4 Distribution of Turf Grass by Sector in Maryland, Virginia and New York 1 Turf Sector MD 2005 VA 2004 NY 2004 Home lawns 82.6% 61.6% 82.1% Apartments 0.6 Nd 0.8 Roadside right of way Nd Municipal Open Space Nd Parks Commercial Nd Schools Golf Course Churches/ Cemeteries Airports/Sod farms) As reported in MDASS (2006), VADACS (2006) and NYASS (2004) nd = no data as the indicated turf sector was not sampled or estimated 3.2 Status of State Phosphorus Fertilizer Legislation Three states in the watershed have enacted phosphorus fertilizer legislation as of 2011 (MD, NY, and VA). Pennsylvania is currently considering legislation, but it has not yet been passed. A common feature in all three states is elimination of phosphorus in lawn maintenance fertilizer products. There are many other elements to each state law, and these are compared in Table 5. Some include a ban on winter fertilization applications, expanded product labeling requirements, and prohibitions on applying fertilizer to impervious surfaces or near water features. Some states also establish a certification process for commercial applicators. Maryland has specific requirements on the maximum individual application of N fertilizer, and a minimum requirement for slow release N formulations. The Panel noted that one of the limitations of the new laws is that they did not allocate funds for expanded education and outreach to make their residents aware of the various nutrient management provisions of their respective laws. 15

16 Table 5 Comparison of Bay State Phosphorus Fertilizer Laws 1 Key Elements MD NY VA 2 Year Enacted/Year Effective 2011/ / /2014 P Ban for Lawn Maintenance Fertilizer Yes Yes Yes Winter Application Ban Yes Yes No Product Labeling Requirements Yes Yes Yes Starter Lawn Exemption Yes Yes Yes Organic/Biosolid Exemption No No Yes Retail Display Requirements No Yes No Prohibit Application on Paved Surfaces Yes Yes No Prohibit Application Near Water Features Yes Yes No No Fertilizer Use as a Deicer Yes No Yes Maximum N Fertilizer Application Yes No No Slow Release N Requirement Yes No No Special Requirements for Applicators Yes No Yes Certification of Commercial Applicators Yes No Yes Enforcement and Fines Yes Yes No 1 DE, DC and WV do not have legislation, while it has been introduced but not passed in PA 2 An amendment was passed to the VA legislation in 2012 to include nitrogen in the urban nutrient management regulations that Department of Conservation and Recreation is charged with developing. Consequently, VA may prescribe more specific practices to reduce nutrient loss in future regulations. 3.3 Trends in Non-Farm Fertilizer Sales in the Bay watershed The Panel examined trends in non-farm fertilizer sales statistics, which are tabulated by each state's agricultural statistics agency, as well as sales data from industry sources. The Panel noted that both sources of fertilizer sales data have weaknesses, and that individual state reports are not consistent with other states (e.g., some rely on tonnage of fertilizer products sold, whereas others supply more detailed data on the actual mass of nitrogen and phosphorus sold). Data on the actual nitrogen content of lawn fertilizer sales appears to be very limited. The Panel only saw official state-derived lawn fertilizer sales data from Delaware and it is not clear whether the other Bay states accurately track lawn fertilizer separately from overall fertilizer sales. Some of the best data on lawn fertilizer sales comes from industry sources, particularly the Scotts MiracleGro Company (SMC, 2011), which is the market leader in sales of lawn fertilizer both homeowners and the lawn care service industry. SMC (2011) reports that there has been a substantial decrease from 2006 to 2010 in the overall amount of nitrogen (33%) and phosphorus (77%) in the lawn fertilizer they have sold in the Bay watershed. Unfortunately, the SMC data is incomplete (because SMC accounts for about 60% percent of total lawn fertilizer sales in the watershed), and is not always consistent with the limited official state data that are available. And it raises a number of questions about differences between states that the panel could not answer. 16

17 With these caveats in mind, the Panel looked at the long term trends in non-farm fertilizer data, with a focus on Maryland. Non-farm fertilizer use increased from about 60,000 tons per year in 1990 to about 200,000 tons in 2004 (MDA, 2005). Since then, non-farm fertilizer sales appear to have stabilized, with some recent industry evidence that they have been dropping in the last few years (SMC. 2010). Felton (2007) developed estimates of the non-farm tonnage of nitrogen sold in Maryland from 1994 to 2004 (see Figure 2). The analysis shows a steady rise through 2000, followed by a drop to mid 1990's levels in the last two reporting years. Insufficient data were available to track long term trends in phosphorus non-farm fertilizer sales. Figure 2 Trends in Farm and Non-Farm N Fertilizer Sales in MD from (source: Felton 2007). The industry data also suggests that there has been substantial reduction in the P content of the lawn fertilizer being sold in the Bay watershed states due to SMC's initiative to phase out P in fertilizer products and in anticipation of the implementation of recent state phosphorus fertilizer legislation (Table 6). This trend is supported by the official state data from Delaware (Table 7), which indicates that in the state as a whole the amount of phosphorus contained in non-farm fertilizer being sold decreased 86 percent from

18 Table 6 Industry Reported Change in P Fertilizer Sales in the Bay States, 2006 to State Percent Millions of Millions of reduction Pounds Pounds Pennsylvania % Maryland % Virginia % Delaware % West Virginia % Total % 1 annual sales data reported by SMC (2011) for non-farm fertilizer sales by state. Scott's currently has a 60% market share, and has committed to a full phase out of P in its fertilizer products by January 1, Analysis performed by Gary Felton, Note that the statistics on P sales are provided for each state as a whole, and NOT the fraction of the state located within the Bay watershed The Scotts data also appears to indicate a decline in the sale of nitrogen in lawn fertilizer from , but this trend did not appear to be as pronounced as the trend in phosphorus. Taken together, the industry and limited official state sales data provided sufficient justification in the judgment of a majority of panel members to support a preliminary credit for a reduction in P application rates in the CBWM, based either on statewide legislation or the fact that P lawn fertilizer sales are declining anyway as a result of industry practice. The Panel concluded that any state-wide nutrient reduction credit must ultimately be defined and verified using more detailed and accurate state non-farm fertilizer statistics in the future. The details of these verification protocols are described in Section 6. Table 7. Change in Non-Farm Sales of Phosphate Fertilizer in Delaware 2006 to 2010 Million lbs Change of P2O % Source: Delaware Department of Agriculture, as Reported in DE Final Phase 2 Watershed Implementation Plan (May, 2012) 18

19 3.4 Derivation of the original CBP-approved rate for urban nutrient management The CBP has had an approved nutrient removal rate for urban nutrient management in effect for nearly 15 years (CBP, 1998, Appendix H). The entire documentation for the rate is provided below:... urban nutrient management leads to a reduction in urban fertilizer applied. Urban nutrient management involves public education (targeting urban/suburban residents and business) to encourage reduction of excessive fertilizer use. The CBP Nutrient Subcommittee Tributary Strategy Workgroup has estimated that urban nutrient management reduces nitrogen loads by 17% and phosphorus loads by 22% No scientific or modeling analysis could be found to support or document the nutrient reduction rates cited above. In addition, the Panel noted that the definition of the UNM was extremely ambiguous and could not be accurately measured, tracked or verified. Therefore, the Panel concluded the existing definition and associated removal rates for the existing CBP-approved UNM practice could not be technically justified. The Panel devised a more specific definition for UNM based on ten core lawn management practices that collectively reduce the risk for nutrient export, and devised a more defensible protocol to estimate the nitrogen and phosphorus reduction credits associated with its implementation. 3.5 How nutrient loads from pervious areas are simulated in the context of the CBWM The Chesapeake Bay Watershed Model (CBWM) simulates nutrient dynamics for a broad range of land uses and land covers throughout the watershed, including urban pervious land. Given the central role of the model in deriving TP and TN reductions associated with various levels of UNM practices, it is helpful to understand how the model currently simulates nutrient pathways, processes and export, with a specific focus on key model assumptions on the response of pervious lands to urban fertilizer inputs, and how the fertilizer inputs are derived. The CBWM uses PQUAL to simulate P dynamics within pervious lands, and AGCHEM to simulate N dynamics. The basic documentation for how the model simulates nutrient loadings and BMP reductions can be found in CBP (1998). The phosphorus simulation is fairly straight forward, and is represented in Figure 3. For each unit of pervious land, the model calculates the flow volume to surface runoff, interflow and groundwater. Atmospheric and fertilizer inputs are then applied, and the P export is defined based on the assumed concentration of phosphate and organic phosphorus for each of the three types of flows. As shown in Figure 3, the CBWM has a 50% sensitivity to P inputs, which basically means that only half of the fertilizer input is available for export (the rest is retained in the soil or by plant uptake). The P concentration factors are initially derived 19

20 from literature and monitoring data, but are refined when the model is calibrated to regional water quality monitoring data. Figure 3: Representation of how P Export is Simulated in PQUAL Module of the CBWM (Shenk, 2012) The nitrogen simulation for pervious lands in CBWM operates in much the same fashion as phosphorus, with the exception that it includes the more complex N cycling process as different N species move through soils and plants and are modified by microorganisms (see Figure 4). Atmospheric deposition and fertilizer are the two primary inputs, and exports are based on flow volumes and N concentrations in surface runoff, interflow and groundwater, respectively. The CBWM tends to be very retentive of fertilizer inputs, although they may be transformed into outputs of organic N under some circumstances. 20

21 Figure 4: Conceptual Diagram Showing How Nitrogen is Simulated for Pervious lands in the AGCHEM module of CBWM (Shenk, 2012) Defining Fertilizer inputs The CBWM utilizes a unit "acre" of pervious land, which receives a "weighted average" fertilizer application rate over the entire watershed (which includes areas that are fertilized and not fertilized). The weighted average fertilization rates are derived from fertilizer behavior surveys, agricultural turf grass statistics and non-farm fertilizer sales estimates, and is documented in CBP (2011). The average annual nitrogen fertilizer input on urban land assumed in the CBWM is 43 lbs N/acre/year or expressed in terms of fertilizer bag label directions, about 1 lb N/1000 sf/yr. The corresponding phosphorus fertilizer input is 1.3 lbs P/acre/year or about 0.03 lb P/1000 sf /yr. In the context of the model, fertilizer "applications" are made over an 80 day period in the spring and the fall. The Panel did some cross-checking and confirmed that these rates were an appropriate representation of the aggregate fertilization inputs for pervious land during the period when the CBWM was calibrated. 21

22 22 Section 4 Review of the Available Science In the last decade, there has been a great deal of research to better understand the nutrient dynamics of turf grass "ecosystems" and their relationship to nutrient loads and downstream water quality. The panel reviewed more than 150 papers and reports on these topics. Several important review papers included Soldat and Petrovic (2008), Felton (2007), Daniels et al (2010) and Guillard (2008). This section describes the key findings from the literature review. 4.1 Review of Phosphorus Dynamics on Urban Lawns There are four potential pathways where P can be exported from urban lawns: 1. Leaching into groundwater (usually minor) 2. Soluble P in surface runoff 3. Sediment bound P in surface runoff 4. Organic matter (i.e., leaves and grass clippings) that reach adjacent impervious cover and are washed into the storm drain system Phosphorus leaching is generally only a concern on shallow, sandy or artificially drained soils, as most P seldom leaches more than three feet through the soil (Daniels et al, 2010). Some urban soils may be saturated with respect to P, either because they have been fertilized for many years and/or because they reflect the legacy of past farming activity. In these conditions, soluble P can leave the soil in surface runoff without sediment (e.g., Maguire and Sims, 2002 and Soldat and Petrovic, 2009). P loss can also occur when phosphorus attached to sediment and organic matter are exported by surface runoff. The potential loss is greatest when turf is dormant and particularly when soils are frozen (Bierman et al, 2010a). Turf grass clippings typically contains 2.0 to 5.0% P in dry matter tissue (Soldat and Petrovic, 2008, Guillard and Dest, 2003). Ray (1997) measured the P content of dead leaves at 1.5% of their dry weight. Soldat et al (2009) notes that P can be released by dead vegetation. Dorney (1986) reported that 9 % of total P in leaves was potentially leachable in 2 hours Various studies have evaluated P losses from fertilized lawns. Shuman (2004) noted that losses sharply increased as the P fertilizer application rate increased, but also noted that a certain amount of P loss was independent of fertilizer application. Soldat and Petrovic (2008) reviewed 12 studies and noted that P losses ranged from less than 1% to as much as 18%, depending on turf grass conditions and fertilizer timing. They found that P loss was greatest when storms occurred shortly after P fertilizer applications. P losses were also strongly related to the runoff volume generated by the lawn. Factors that increase runoff volume (e.g., steep slopes, compacted soils, frozen ground, low turf density) are all associated with a higher risk of P loss.

23 4.2 Review of Nitrogen Dynamics on Urban Lawns There are four primary sources of nitrogen to urban lawns: mineralization of N in the soil, atmospheric deposition, degradation of organic matter (such as lawn clippings) and fertilizer inputs. While the rates of soil mineralization are very site-dependent, there are good data on atmospheric deposition rates. Measured atmospheric deposition in Baltimore was 0.23 lbs N 1000 sf/year (Groffman et al, 2011) which is generally consistent with the Baywide average N deposition of 0.42 lbs N 1000 sf/year which is the current average input to pervious lands in CBWM. Decomposition of lawn clippings is another important source of N to the lawn, as they rapidly become available in the soil (Raciti et al, 2011a). Frank et al (2005), Felton (2007) and Kopp and Guillard (2004) independently estimated that returning grass clippings to the lawn could provide approximately one lb of N/1000 sf/year. Estimates for average fertilizer applications are provided in Sections 3.3 and 3.5. There are four potential pathways where N can be exported from urban lawns: 1. Leaching of nitrate into groundwater 2. Loss of nitrate and ammonium in overland flow 3. Organic nitrogen (e.g., lawn clippings or N attached to eroded sediments that runs off or is blown over to adjacent impervious cover and is washed into the storm drain system, and 4. Volatilization of ammonia into the atmosphere shortly after fertilization Nitrate Leaching Nitrate leaching can be a significant source of N export under certain lawn conditions, and is dependent on soil type, irrigation, grass species, rooting depth and fertilization rate and timing (Bowman et al 2002, and Pare et al 2006). Nitrate leaching is greatest during the seasons of the year when the grass is dormant. Cool season turf grass typically goes dormant sometime in December and resumes growth at some point in February or March, depending on the severity of the winter. Cool season turf grass may also go dormant in the summer due to extensive drought or heat. The measured N loss via leaching is related to the amount of water soluble fertilizer applied. Table 8 presents the results from 16 different research treatments that measured TN or nitrate loss as a function of fertilization rate/frequency. The analysis indicates relatively low N losses for lawns that applied less than 130 lbs N/yr (or >3 lbs N per 1000 sf lawn; shaded in green. By contrast, N losses were significant higher for lawns with N fertilizer treatments that exceeded the 3 lb threshold (shaded in red in Table 8). N losses were also influenced by the type of fertilizer and the number of soluble N applications. 23

24 Table 8: N Losses from Turf Grass as a Function of Fertilizer Application Rate N Load N Fertilizer % of Notes Exported (lb/ac) Input (lb/ac) Fertilizer Exported 1 Reference % Mancino & Troll, 1990 In 10 weekly apps % Namcino & Troll, 1990 In 5 biweekly apps % Spence et al High Maintenance Fescue lawn % Spence et al 2012 Low Maintenance Fescue Lawn % Frank et al Lo input leaching losses % Guillard & Kopp 2004 Organic fertilizer % Mancino & Troll, 1990 Single application % Guillard & Kopp, 2004 PCSCU slow release % Quiroga-Garza et al. Semi-arid, Warm season 2001 Bermuda grass % Erickson 2001 Leaching loss % King et al Hi Risk: Watered to maintain 85% FC with tile drains Hi Risk: Hi Input semi-arid Bermuda grass % Quiroga-Garza et al % Guillard & Kopp 2004 Hi Risk: Highly soluble ammonium nitrate % Frank et al Hi Risk: Hi Input % Roy et al 2000 Hi Risk: 3x sod grower practice overwhelms turf, fall leaching losses %-71% Pare et al 2006 Hi Risk: 80:20 sand peat media, applied 25kg/ha biweekly over 7 month growing season. Multiple cultivars. 1 Export is calculated as % fertilizer inputs. This overestimates turf system exports for field studies with atmospheric inputs in precipitation. Not all studies measured all species of nitrogen, and some may have measured only surface or subsurface N losses Historically, concerns with nitrogen leaching from lawns have been driven by human health concerns regarding nitrate contamination of drinking water particularly groundwater supplies. For this reason the concentration of leachate remains a significant concern. When it comes to urban nitrogen load reduction, however, nitrate leaching are not synonymous with total N loads delivered to the Bay. Nitrate leaching introduces soluble nitrogen into subsurface flow paths that may encounter reducing conditions supporting denitrification. Indeed, the potential for denitrification along subsurface flow paths is a principal nitrogen removal mechanism expected from riparian and vegetated buffers. Although leaching losses are not equivalent to surface losses, nitrate leaching in landscapes with 24

25 highly permeable soils and high water tables pose the greatest risk for transforming leachate into surface loads through shallow subsurface return flows. Recent research indicates that lawns are highly retentive of fertilizer N under typical application rates and lawn conditions. Groffman et al (2004) found approx 75% of fertilizer N was retained in urban lawns monitored in Baltimore. Kaushal et al (2011) used N isotopic ratio signatures to show watershed export of nitrogen is not directly proportional to fertilizer inputs in Baltimore watersheds. Though lawn fertilizer is a significant input to the watersheds, the isotopic signatures of stream nitrogen suggest sewage is a much more significant N loading source than lawn fertilizer. Raciti et al (2008) and Raciti et al (2011b) demonstrated residential lawns have a high capacity for both carbon and nitrogen storage in plant biomass, thatch and soils. Denitrification in fertilized urban soils is significant at certain times of the year, with a loss up to 0.30 lbs/1000 sf/year, nearly all of which occurred during less than 5% of the growing season when soils are saturated and air temperatures are warm (Raciti et al, 2011a). A lawn's capacity for N storage and transient seasonal conditions supporting high de-nitrification rates may explain why other research studies found relatively low N export, despite significant N fertilizer inputs. Nitrate loss in Overland Flow A recent study measured nitrate-n losses in overland flow over 87 rainfall events from low and high maintenance lawns in the North Carolina piedmont (Spence et al, 2012). The authors found that the highly maintained lawns (fertilizer, irrigation and reseeding) generated slightly less runoff (runoff coefficient, Rv= 0.04) and nitrogen export (about 1% of N fertilization applied) than lawns with a less intense maintenance regime (which still included fertilization). The less maintained lawns had a Rv of 0.06 and produced runoff during more rainfall events and generated slightly higher yields of nitrate, compared to the high maintenance lawns. The authors did note that their test lawns were located on undisturbed and highly permeable soils, which may not be representative of all residential situations. The Panel concluded that several risk factors sharply increased the risk of overland flow and potential fertilizer export. The amount of runoff volume is largely determined by lawn slope, soil compaction, and turf density. For example, Garn (2002) found that runoff was as much as 50% greater in steeply sloping urban lawns. Runoff losses appear greatest during the seasons of the year when the grass is dormant or the ground is either saturated or frozen (Guillard et al, 2008). Easton and Petrovic (2008) noted that N losses were greatest in newly established turf. N loss was most closely associated with shallow and compacted soils that had low water storage capacity. Loss of Organic N in Surface Runoff. Another N export pathway involves the loss of organic nitrogen in surface runoff. The organic nitrogen may be derived from lawn clippings, leaves and eroded sediments that are blown or washed off lawns and into the storm drain system. Several authors have 25

26 indicated that this may be an important N export mechanism (Daniels et al, 2010 and Felton, 2007) given the rapid rate of decomposition and release of lawn organic matter. Spence et al (2012) note that the N content of lawn clippings ranged from 2.7 to 4.5% of their dry weight. Source area sampling of lawn runoff by Steuer et al (1997) measured a median TN concentration of 9.7 mg/l, 90% of which was measured as TKN. Lawn N concentrations were more than four times higher than N concentration in streets, parking lots and rooftops sampled in the same study. Other researchers have also show that organic forms of nitrogen predominate over nitrate in lawn runoff (Garn, 2002, Spence et al, 2012). While significant concentrations of particulate organic N have been measured in lawn runoff, the significance of this loss pathway is less clear when it comes the total N export. For example, the high particulate organic N loads reported by Garn (2002) were attributed to leaf litter, rather than grass clippings. While the particulate N concentrations for suburban lawns sampled by Spence et al (2012) were high, the total particulate N load exported was less than 0.15 lbs/ac/yr, regardless of lawn maintenance regime. Volatilization Some organic forms of fertilizer, especially urea, may be subject to volatilization losses shortly after they are applied. The organic fertilizer may be converted to ammonia which can be lost to the atmosphere. Volatilization occurs on warm and moist soils, and can be reduced if fertilizer is watered in immediately after application (Felton, 20o7). In summary, while lawns have been shown to be retentive of fertilizer nitrogen under most conditions, they can produce significant N losses via leaching, runoff, and clippings in high risk conditions (see next section for a detailed list). 4.3 High Risk Nutrient Export Factors. The Panel noted that lawn nutrient export was a classic example of the "disproportionality" concept cited by Baker et al (2008). The basic concept is that most lawns in the urban landscape are reasonably retentive of nutrients under most conditions, with a small proportion of high risk lawn conditions or behaviors responsible for most of the total nutrient export. Baker et al (2008) argue that an UNM program that is specifically targeted to high risk lawns would be the most effective, economical and fair as it would focus on lawns that provide the greatest source loading. 26

27 Application Rate Rate None None Recommended Excessive Low Slope/Soil Risk High Flat to gentle slopes Healthy well- structured soil Application rates > 6 lbs/k High wear turf Application rates 3-6 lbs/k Extension Recommendation Application rates 2-3 lbs/k Low input Application rates > 1-2 lbs/k Zero Input Residential Commercial Institutional Public parks/rec Public ROW Steep poorly drained slopes or Sandy well drained soils with shallow water table Soil test (including ph) Mulch Clippings Multiple small well-timed applications net clippings No fertilizer or clippings on pavement Weather sensitive application 1-2 applications as convenient Some attention to convenient recommendations Broadcast WS fertilizer on schedule, early spring, late fall and then some just to be sure Remove clippings Practices Good Poor Figure 5 Conceptual Model for Defining N Export Risk in the Urban Landscape (developed by Stuart Schwartz) A range of landscape and behavioral factors affect the relative risk of nutrient loss and therefore the effectiveness of urban nutrient management (UNM) from turf grass land uses. The nutrient loading risk from turf grass in any distinct urban land use (residential, commercial, institutional, etc.) may vary due to the slopes and soils, the fertilizer application rate adopted, and the quality of the lawn care practices being employed (see Figure 5). Slope/Soil Risk: For any land use, steeper slopes will tend to increase the risk of runoff and therefore surface transport of sediment and nutrients. The slope risk interacts with the soil texture and structure. Thick loamy soils on gentle slopes have a very low runoff loading risk. Poorly drained soils on steep slopes produce high runoff, and hence, a higher risk for nutrient and sediment loading. In other cases, sandy, well-drained soils in areas with shallow water tables may also present a high risk of transporting dissolved nutrients mobilized through leaching, that may return to surface water through shallow subsurface flow paths. Fertilizer Application Rate: Nutrient loading risk is further compounded by the nutrient application rate employed. Across each land use/slope-soil risk category shown in Figure 5, land managers may elect to apply widely different fertilization rates, ranging from zero to application rates in excess of 6 lbs/1,000 sf that would be considered 27

28 excessive for normal or high wear turf. Between these extremes land mangers and homeowners may elect low input lawns applying 1-2 lbs/1,000 sf; maximum extension recommended rate of 3 lbs/1000 sf; or high intensity fertilization of 3-5 lbs/1,ooo sf that are sometimes suggested for heavily stressed turf such as athletic fields. Lawn Care Practices: Finally, the nutrient loading risk suggested by the convolution of land use, slope-soil risk, and fertilization application is further refined by the type of the overall lawn care practices employed. For example, the ten core UNM practices recommended by the Panel should tend to minimize the risk of N export, and to a lesser degree, P export. By contrast, high risk lawn care practices may involve broadcasting water soluble fertilizer on a routine schedule irrespective of weather or turf conditions, and then adding a little more, because more must be better. Grass clippings are removed rather than recycled on the lawn, and an extra application of fertilizer is applied in late fall or even early winter, to jump start spring "greening". Between these extremes of low and high risk practices exist a continuum of moderate risk practices. These lawns may implement some, but not all of the recommended UNM practices (e.g., not closely coordinating application timing and irrigation). Some of the recommended lawn care practices may be incorporated informally (e.g. multiple fertilizer applications) as convenient by the homeowner without having a written UNM plan. The spectrum of possible lawn care practices may further moderate or amplify the risk of nutrient export. Together, these three major dimensions of risk associated with turf grass fertilizer use -- landscape factors, fertilizer application rate and lawn care practice -- interact to affect nutrient export from urban pervious land to the Bay. The current CBWM, however, is limited to a single, generic urban pervious land use and does not consider the heterogeneity of turf grass based on those risks. The Panel considered these model limitations and attempted to account for a risk-based approach to define UNM credits. The Panel concurred with the targeting approach, and reviewed the literature to define a more operational definition of what constitutes high risk conditions or behaviors. They include lawns with: Owners are currently over-fertilizing beyond state or extension recommendations 2. P-saturated soils as determined by a soil analysis 3. Newly established turf (Easton and Petrovic, 2004, Line and White, 2007) 4. Steep slopes (more than 15%) 5. Exposed soil (more than 5 % for managed turf and 15% for unmanaged turf) 6. High water table (within three feet of surface ) 7. Over-irrigated lawns (Barton and Colmer, 2005, Guillard, 2008) 8. Soils that are shallow, compacted or low water holding capacity (Easton and Petrovic 2008a and b) 9. High use areas (e.g., athletic fields, golf courses)

29 10. Sandy soils (infiltration rate more than 2 inches per hour) 11. Adjacent to stream, river or Bay (within 300 feet) 12. Karst terrain UNM planning agencies may elect to identify additional factors to define high risk lawns; a list of environmentally sensitive factors such as those defined in Virginia's Nutrient Management Standards are provided in Table 9. Some of the high risk factors could be mapped or measured at the local level using available GIS data, neighborhood and/or site surveys or soil sample analysis. The Panel recommends that planners screen for high risk factors when developing individual UNM plans and designing community outreach programs. The Panel also recommends higher UNM nutrient reduction credits be granted when effective targeting based on high risk factors and behavior change can be confirmed and verified. Table 9. Additional Virginia UNM High Risk Factors Stipulated by Regulation "Environmentally sensitive site" means any pervious land which is particularly susceptible to nutrient loss to groundwater or surface water since it contains, or drains to areas which contain, sinkholes, or where at least 33% of the area in a specific field contains one or any combination of the following features: 1. Soils with high potential for leaching based on soil texture or excessive drainage 2. Shallow soils less than 41 inches deep likely to be located over fractured or limestone bedrock 3. Subsurface tile drains 4. Soils with high potential for subsurface lateral flow based on soil texture and poor drainage 5. Floodplains as identified by soils prone to frequent flooding in county soil surveys 6. Lands with slopes greater than 15%. Source: VA DCR (2005) 4.4 Scientific Justification for Core UNM Practices The Panel focused considerable efforts to define ten specific lawn care practices that are most strongly associated with reduced nutrient export from turf grass areas. The Panel primarily focused on practices that could reduce nitrogen export, given the effect of state phosphorus fertilizer legislation and the recent industry phase out of phosphorus in fertilizer products. However, several of the lawn care practices employed to reduce nitrogen loss also have the potential to reduce phosphorus loss. The scientific justification for these core practices are described in this section. The Panel acknowledged that each Bay state should adapt and modify these recommendations to reflect their unique conditions, as well as the recommendations of state lawn care extension agencies. Specific elements of the core UNM practices may 29

30 differ across in the watershed, especially with respect to warm or cool season grass species and different climatic or plant hardiness zones. Lawn Care Practice 1. Consult with the local extension service office, certified plan writer or applicator to get technical assistance to develop an effective urban nutrient management plan for the property, based on a soil test analysis. The precise lawn care prescription should be based on state-specific UNM recommendations or regulations, as well as an understanding of soil properties, the type of grass species, the age of the lawn, and other factors. Professional expertise is essential to develop an effective plan. Lawn Care Practice 2. Maintain a dense vegetative cover of turf grass and/or conservation landscaping to reduce runoff, prevent erosion, and retain nutrients The research demonstrates that dense vegetative cover helps to reduce surface runoff which can be responsible for significant nutrient export from the lawn, regardless of whether it is fertilized or not. Dense cover has been shown to reduce surface runoff volumes in a wide range of geographic settings and soil conditions (Easton and Petrovic, 2004, 2008a,b, Garn, 2002, Bierman et al 2010, Ohno et al, 2007, Raciti et al, 2008, Shuman, 2004, Vlach et al, 2008, Legg et al, 1996 and Spence et al, 2012). If a lawn does not have a dense cover, it has an elevated risk for nutrient export, especially if soils are compacted or slopes are steep. In these situations, the primary nutrient management practice is to identify the factors responsible for the poor turf cover, and implement practices to improve it (e.g., tilling, soil amendments, fertilization or conservation landscaping). Lawn Care Practice 3. Per the UNM plan, Choose not to fertilize, OR Adopt a Reduce Rate/Monitor Strategy, OR Apply less than a pound of N per 1000 square feet per each individual application. The Panel noted that three distinct and acceptable N fertilization strategies exist to effectively reduce the risk of export in runoff or via leaching, depending on site conditions and the needs and preferences of the homeowner. 30 The first strategy is to elect to not fertilize at all, which may be appropriate for relatively flat, mature lawns with a dense vegetative cover (e.g., older than ten years). This strategy relies on soil mineralization, lawn clippings and atmospheric deposition to supply the N inputs needed for growth, and is effective as long as turf cover remains dense (see Practice 2). (Caution: this strategy should not be employed on lawns that have poor turf cover or exposed soils since their runoff has a higher risk of phosphorus and sediment export, according to research. The second strategy utilizes a "reduced rate and monitor" approach to fertilization advocated by Guillard et al (2008). In this strategy, the homeowner reduces application rates on the fertilizer bag label by one-third to a half and

31 monitors the lawn response over time. The homeowner only re-applies fertilizer (at the smaller dose) if they perceive that lawn quality starts to fall below acceptable levels. Consumer research shows that most residents follow fertilizer label information to decide how much to apply (Schueler, 2000, Kerr and Downs Research, 2011), so that this iterative approach to lawn management could be effective. The third strategy is to fertilize at the state or cooperative extension recommended N fertilization rate but split it into 3 or 4 small doses during the growing season. In MD and NJ, this recommended rate is defined as a maximum single application of no more than 0.9 pound of N per 1000 square feet; other states and/or extension recommendation in the watershed may be slightly different. This strategy greatly reduces the N export risk for homeowners that desire a green lawn or use a lawn care company. Several studies provide strong evidence for the second and third strategies, i.e., that it is better from a water quality perspective to apply smaller doses several times a year rather than the single maximum dose. Frank et al (2006) demonstrated the smaller dose strategy reduced N export for mature Kentucky bluegrass turf. Easton and Petrovic (2004) reported reduced P loss in leachate and runoff from a sandy loam soil when the same annual fertilizer application rate was spread over four smaller applications rather than two larger ones. Daniels et al (2010) also recommends the small dose fertilizer strategy for the Commonwealth of Virginia. The "choose not to fertilize" option should not be used if the lawn has poor turf cover...these un-managed lawns can deliver runoff, sediment and nutrients to the stream network 31